We usually disable zebra. Something like a route server.
Not sure if this hits other folks as a default behavior, but...
we get thousands of warning messages like:
sendmsg_nexthop: Can't send NH register, Zebra client not established
Recommended approach on this one?
John Kemp
---
Hi Timo,
Thanks. I'll try get this in the next slurpage. Due to travel that might
be after a week on tuesday though.
Aim will be to have the "staging/proposed -> push to master" cycle to 1
to max 2 weeks here-after.
(Modulo the odd cases of people saying "hey, I'm away that week or two
-
The FIB override routes can override ZEBRA_FLAG_SELECTED routes
in FIB. Use the FIB state instead to report correct nexthop when
FIB override routes are present.
---
zebra/zebra_rnh.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/zebra/zebra_rnh.c b/zebra/zebra_rnh.c
index
Hi,
Now that we have all the important stuff in master (fib overrides,
nexthop tracking), I'm ready to submit NHRP for inclusion.
The first three commits fix simple bugs. The first two ones probably
are not triggerable without running NHRP. The third commit may have
caused regressions for some
This reverts commit 7e3a435bd99 "A valid BGP nexthop is flagged as invalid"
Problem is BGP thinks the nexthop is accessible when it's recursive, and
selects it, but zebra rejects it at route install time. Causing FIB and
BGP state to be out-of-sync. Fix nht to follow same rules as zebra rib.
---
Philippe,
I think either we're talking past each other are have some other
major disconnect.
Per BGP LxVPNs, we really should distinguish between VRFs and (vpn)
route distribution.
WRT VRFs:
The contents of a VRF imported from BGP is determined by one or more
(import) RTs. RDs are ignored
On Tue, Oct 18, 2016 at 3:56 PM, Lou Berger wrote:
Hello Lou,
Please read below my comments.
Consecutive to your feedback, I think there are some improvements to bring,
mainly about configuration and structure naming.
> It looks like RD is your key to a VRF, is this right?
>